TPS1641: TPS16411 Auto-Retry ?

Hi Martin,

Thanks for reaching out!

The description in the table is correct. What is the test condition? Can you share test waveforms of Vin, Vout, input current and FLTb to review.

Best Regards,

Rakesh

Hi Rakesh,
my application needs 24V
1.in the EVM kit doc shema the input voltage is described with 10.8 to 28V

i think this was my first mistake that i thought 24V is ok. Startup with S2 released is fine. But pressing and releasing again is not working with 24V.
I have to decrease the input voltage below ~21V. So it is clear for me why no autoretry didn't work. Now TPS16410 and TPS16411 has the same recovery if PLim is to high.

2. i made the same tests with TPS16411 (replaced in U2). VIN2 is connected to J9-1. Q1 bypassed. There is the same auto-recovery behaviour as TPS16410 if Plim is to high or if i short vout2 to gnd. No latch off.
So "latch off" is only for thermal shutdown and short vout to vin ?
With multiple shorts from vin to vout with breaks I can switch on Plim again but Flt will stay low. is this an expected behaviour ?

best regards

Martin

Hi Martin,

I will get back before thursday on this.

Regards

Kunal Goel

Hi Martin,

1. Regarding channel 1 if you will replace FLT pullup resistor with high value like 100K you will see it working at 24V even when switch is pressed and released.

2. Now regarding channel 2, 

what is the load type used?

It is not looking that device is retrying because retry interval should be less. How much PDLY you have chosen? Retry time should be 8*tpdly. 

Can you tell all the jumper positions you used on channel 2 ?

Also please share device marking.

Can you share EVM image also?

Regards

Kunal Goel

Hi,

C10 changed to 22nF (now same as C2)

1. Test yesterday was with this electronic load - mode CR (second picture)

2. Test today with old fashioned resistive load

with this load blanking time is 54ms - ok, power limit duration 108ms -ok

auto retry 8x = 433ms ok, so electronic load is not good for testing 

I think this is ok.

So "latch off" with TPS16411 is only for thermal shutdown and short vout to vin (working with real resistive load and stronger power supply) ?

Best regards,

Martin

Hi Martin,

1 . For any eFuse we do not recommend to use CC load while performing current/power limiting test. Because eFuse in current limit and cc load will be like two different values of current sources in series which will give unexpected results.

2. Is TPS16411 retrying indefinitely? It should latch even if FET is turned of because of power limit duration expiry. Let me check on this. 

Regards

Kunal Goel

Hi Kunal, 

i tested now with my real hardware load (input voltage could be 10V-26V, max. 1A, a TPS26600 with LMR23630 and TPS54061), with both TPS1641x on the EVM, and after a few seconds fuse will shutdown. I can not see that a limit has been exceeded. Vin,Vout and EN seems to be ok. The current ripple is about 100mA and f=2kHz. If i do the test with a electronic load switching between 0.25A and 0.3A with a higher frequency (~100Hz) i see the same behavoiur. (Rplim=255K, Riocp=7K32)

Do you have an idea ? With the output capacitor it is a little bit better but didn't solve the problem. Do we need C6 (EVM DNP) ? 
Various delays are also not a solution.

best regards,
Martin

with 24V:

Startup with 12V, no retry ? 

Hi Martin,

Lets proceed step by step. Lets first debug why latch variant was retrying. It should latch.

Can you place TPS16411 on channel 1 and repeat power limit test? Let see if it retries on both channel.

Regards

Kunal Goel

Hi Kunal,
The desoldering could be done next Tuesday at the earliest. Actually I would like to carry out the further tests with the original component before I change this. Would that be possible ? 
best regards,

Martin

Hi Martin,

Can you try increasing load switching frequency beyond 15khz? I think it could be related to that. At higher frequency frequency we should not see shutdown.

Regards

Kunal Goel

Hi Kunal,

with load changes from 0.2A to 0.3A  at VIN1=12V (EVAL TPS16410, ILimit 2.23A, PLimit=38W) the efuse does a shutdown at load change frequency between ~50Hz and ~4kHz. Lower and higher frequencies are working in this testcase.  

The problem is that our load is an already finished intelligent terminal (processor, display, interfaces) and we have no influence on the frequencies, variations and strength of the current consumption.

I think additional filters in this frequency ranges are no solution.
I tested with more ceramic capacity direct onboard beside D1 (2x10uF/50V) and with no C2. No significant change.

With p-spice simulation (small changes in std. example TPS16140=1 R6=255K, R9=60, ROFF=120, and eg. PW=300u, PER=600u) we do not see this behavior.
Layout in datasheet is nearly the same as layout in EVM with U1. 

Do you have a suggestion ? (in the moment it doesn't matter if latch-off or auto-retry will be used)

best regards,
Martin 

Hi Martin,

Let me discuss with team and get back on this.

Regards

Kunal Goel

Hi Martin,

I will answer you over e2e messaging. Please accept my friendship request.

Regards

Kunal Goel